Menu

AI

Everything You Need to Know About Salesforce SIforce

Key Takeaways:

  • Salesforce SIforce, formerly AIforce, launched at Dreamforce 2026, brings Salesforce data, workflows, and permissions to Claude, Slack, and Lightning.
  • It is an interface layer, not a rebrand: Agentforce supplies the agents, and SIforce carries them to where people work.
  • It launched with Claudeforce, Slackforce, and Agentforce Coworker, with more interfaces on the way.
  • Every request runs on existing permissions and business rules, with Zero Data Retention and no migration.
  • Pricing is unpublished, so teams should audit access, clean their data, and pilot before scaling.

Ever wished you could update a Salesforce record, review a deal, or kick off a workflow without opening Salesforce? Meet Salesforce SIforce, the live interface layer Salesforce unveiled at Dreamforce on September 15, 20261. Launched as AIforce and renamed as SIforce, It carries the data, workflows, business logic, and permissions inside your org to the AI tools your teams already use, such as Claude, Slack, and Lightning.

The idea is simple: your people stop going to Salesforce, and Salesforce comes to them. In this blog, we’ll look at what Salesforce SIforce is, how it differs from Agentforce, what launched with it, and how to get your org ready. Let’s dive right in.

Salesforce SIforce

What Is Salesforce SIforce?

Salesforce SIforce is a live interface layer that brings Salesforce data, workflows, business logic, permissions, and governance to any AI interface. Instead of logging in and clicking through layouts and fields, people and AI agents reach Salesforce from wherever they already work.

Salesforce’s own reasoning is that most company knowledge already sits in Salesforce, but people could only reach it by opening the app. That limited who used it, how much information they could see at once, and how much value they got from it, according to the SIforce announcement2.

Before and After SIforce

Nothing about your data, permissions, or business rules moves. What changes is where people reach them.

Here’s an example from the launch. A sales manager asks Slackbot which accounts are going quiet. Slackbot reads the support cases and Slack threads behind those accounts, explains why, reassigns an owner, creates a follow-up task, and drafts a win-back email, all without anyone opening Salesforce.

Salesforce describes five properties that SIforce adds to any AI interface:

  • Intelligent and Dynamic: Agents reason across the data, logic, and conversations in Salesforce at once, so users get insights instead of lists of records.
  • Secure and Governed: Every request runs on your existing permissions and business rules. An agent sees only what the person asking can see, and every action routes back through Salesforce.
  • Composable: Anyone can build an interface, agent, or view by describing it. Teams that want to go further get MCPs, APIs, plug-ins, and skills.
  • Zero Data Retention: Business data answers the question at hand and is not retained by the model provider.
  • Works with What You Have: There is no new permissions model, no migration, and no custom integration. Admins connect once.

On pricing, Salesforce has not published SIforce pricing yet. The announcement notes that pricing and packaging are subject to change and that availability may vary by region.

Is Salesforce SIforce Different from Agentforce?

Yes. Agentforce and SIforce are separate layers that work together, and SIforce is not a rebrand the way Agentforce replaced Einstein Copilot.

Agentforce is Salesforce’s digital workforce: ready-to-deploy agents plus a builder for making your own. SIforce is the layer that carries Salesforce, including those agents, to the interfaces where people already work. If you have built Agentforce agents, that work carries over, because Agentforce Coworker calls the specialized agents you have already deployed.

Agentforce SIforce

What it is

A digital workforce of AI agents

A live interface layer

What it does

Agents handle tasks for customers and employees

Brings Salesforce data, logic, permissions, and agents to any AI interface

Where people use it

The channels where you deploy your agents

Claude, Slack, Lightning, and more surfaces to come

What you build

Agents, using Agentforce Builder

Interfaces and views, by describing them

In short, Agentforce decides what work gets done. SIforce decides where people and agents can reach Salesforce to do it.

Get Your Salesforce Org SIforce-Ready.

The Backstory of Salesforce SIforce

SIforce did not appear overnight. Marc Benioff previewed it on Salesforce’s second-quarter earnings call in August 2026, calling it an enterprise harness that had evolved over three years, Constellation Research3 reported. The same week, Salesforce and Anthropic announced Claudeforce, the first integration built on it.

The timing matters. Earlier in 2026, enterprise software stocks fell sharply on fears that AI would make SaaS obsolete, a CNBC report cited by The Next Web4 noted. Opening the Dreamforce keynote, Benioff dismissed the so-called SaaSpocalypse and argued it marks the end of software that makes humans do all the work, not the end of software.

SIforce is Salesforce’s answer to that pressure. Constellation Research described it as a way to connect enterprise applications to the very AI labs that were expected to sink SaaS. Benioff framed it as the next interface shift, following the moves from DOS to graphical interfaces, then to the web, then to mobile.

How Salesforce SIforce Works

Salesforce SIforce sits on top of three layers Salesforce already sells and carries them to the tools people use. Data 360 supplies unified context, Customer 360 supplies business logic and permissions, and Agentforce supplies the agents. SIforce exposes all three through the Headless Toolkit, an open architecture of MCPs, APIs, plug-ins, and skills.

Salesforce SIforce Architecture

Every request from Claude, Slack, or Lightning passes through SIforce, which applies your existing permissions and business rules before anything reaches the platform.

What Launched with Salesforce SIforce?

SIforce launched with three interfaces: Claudeforce, Slackforce, and Agentforce Coworker. Salesforce says more will be announced soon. Here’s a closer look at each.

1. Claudeforce

Claudeforce extends the partnership between Salesforce and Anthropic. Its first product, Salesforce in Claude, packages Salesforce into a prebuilt MCP server inside Claude, so teams skip manual setup, complex authentication, and custom skill mapping.

  • It ships with 37 prebuilt sales skills, covering work from prospecting to pipeline hygiene.
  • Deloitte, GitLab, and Legora piloted it, and it is now in beta for all customers. Customers request access on AgentExchange with their org ID, according to The Next Web5.
  • Salesforce plans to add Tableau analytics plus skills for service, marketing, commerce, and industries.
  • For developers, the Salesforce Development plug-in for Claude Code adds more than 40 skills.

2. Slackforce

Slackforce brings Salesforce context into Slack conversations and workflows. It has four parts:

  • Slackforce Surfaces: Live, interactive views built from Salesforce, Slack, and other tools that a team can filter, explore, comment on, and act on together.
  • Slackbot: A personal assistant that reasons across Slack conversations and Salesforce data, then takes governed actions.
  • Slack CRM: Create accounts, log call notes, and update records with a prompt, right from Slack.
  • Slack Code: A shared Slack space where a team works with a coding agent on a project, producing working code, prototypes, and documents.

3. Agentforce Coworker

Agentforce Coworker is an AI teammate inside the Lightning interface. It reasons across accounts, activity, and history, then surfaces insights and takes action.

  • It runs inside Salesforce, within your existing permissions, and nothing is retained outside your trust boundary.
  • It calls the specialized Agentforce agents you have already built, so it grows with your agent workforce.
  • It is available to Salesforce customers now and switches on with the push of a button, with no migration and no new permissions model.

What Can Your Teams Do with Salesforce SIforce?

SIforce widens who can use Salesforce. People who have never opened a dashboard can ask questions, update records, and trigger workflows from tools they already know. Here’s how that plays out by role:

  • Sales reps: Work deals from inside Claude, using prebuilt skills that run from prospecting to pipeline hygiene, with every update written back to Salesforce.
  • Sales managers: Describe a pipeline view in Slack and get a live, shared interface the team can filter and act on, instead of waiting for someone to build a report.
  • Service and account teams: Ask Agentforce Coworker in Lightning to reason across an account’s activity and history, then take the next action.
  • Admins: Connect SIforce once and keep the permissions model you already maintain, since every request respects existing access and business rules.
  • Developers: Build on the Headless Toolkit’s MCPs, APIs, plug-ins, and skills, or code with agents in Claude Code and Slack Code.

Agents gain the most. They can read hundreds of records at once, pull in connected systems, and act with a full view of the business, which no person clicking through screens can match.

How to Get Your Salesforce Org Ready for SIforce

SIforce runs on the permissions, data, and business rules you already have. That is its biggest strength and its biggest risk: whatever is messy in your org now travels to Claude and Slack too. Six steps will put you in good shape:

  • 1. Audit permissions and sharing rules. SIforce answers with whatever the person asking can see. Over-broad profiles and permission sets will now surface data in more places than Lightning.
  • 2. Clean the data agents will read. Agents reason across hundreds of records at once, so duplicates, stale fields, and missing values show up directly in their answers.
  • 3. Tighten your business rules. Every action routes back through Salesforce flows, validation rules, and approvals. Gaps in that logic become gaps in what agents do.
  • 4. Start with one surface and one team. Agentforce Coworker is available now in Lightning, while Salesforce in Claude is in beta. A focused pilot shows where your org needs work before a wider rollout.
  • 5. Set rules for user-built interfaces. Anyone can now describe a view into existence. Decide who can share those views and how they are reviewed.
  • 6. Review your existing Agentforce agents. Coworker calls the agents you have already built, so their instructions and actions will reach more users.

What the Future Holds for Salesforce SIforce

Salesforce has signaled that SIforce will keep expanding. More interfaces are on the way, and Salesforce in Claude will add Tableau analytics plus skills for service, marketing, commerce, and industries.

On stage, Benioff also promised an SDK for building SIforce apps and an SIforce Max Edition that bundles everything into one product, The Next Web6 reported. Meanwhile, the AgentExchange ecosystem is building on the same Headless Toolkit, with interfaces from Anthropic, AWS, Google, and Microsoft and tools from companies like Docusign, Gamma, Jasper, and Rippling.

Two questions remain open: what SIforce will cost, and when Salesforce in Claude moves from beta to general availability. We will update this post as Salesforce shares details.

That’s a Wrap

SIforce changes where Salesforce work happens, not what Salesforce is. Your data, logic, and permissions still do the heavy lifting; they now reach people in Claude, Slack, and Lightning.

The teams that benefit first will be the ones whose orgs are clean and well governed. If you want help getting there, explore Achieva’s Agentforce services and Slack integration services.

References:

Frequently Asked Questions

Salesforce SIforce is a live interface layer, launched at Dreamforce on September 15, 2026. It brings Salesforce data, workflows, business logic, permissions, and governance to AI interfaces such as Claude, Slack, and Lightning.

No. Agentforce is Salesforce's workforce of AI agents. SIforce is the layer that carries Salesforce, including those agents, to wherever people work.

Not entirely. Lightning remains one of the surfaces, with Agentforce Coworker running inside it. SIforce means people no longer have to open Salesforce to get work done.

Claudeforce is the Salesforce and Anthropic partnership that brings Salesforce into Claude. Its first product, Salesforce in Claude, is a prebuilt MCP server with 37 sales skills, now in beta for all customers.

SIforce is built with Zero Data Retention. According to Salesforce, business data answers the question at hand and is not retained by the model provider.

Latest Blogs

Read All >
Salesforce Einstein Analytics and Tableau: CRM Analytics, BI, and AI Forecasting Explained

Salesforce Einstein Analytics and Tableau: CRM Analytics, BI, and AI Forecasting Explained

Key Takeaways: Salesforce Einstein Analytics was renamed, not retired, and is now CRM Analytics. The...

Responsible AI in Salesforce CRM: Governance, Hallucination Risks and Explainability

Responsible AI in Salesforce CRM: Governance, Hallucination Risks and Explainability

Executive Summary AI in Salesforce requires shared responsibility across Salesforce controls and customer governance. Higher...

Transforming Sales with Enterprise AI: Breaking Down Salesforce’s xGen-Sales and xLAM Models

Transforming Sales with Enterprise AI: Breaking Down Salesforce’s xGen-Sales and xLAM Models

Sales teams spend nearly two-thirds of their time on tasks that don’t bring in customers...

Leverage Cloud, Grow Faster.

Explore New Possibilities with Salesforce.

We are Salesforce Summit partner,
taking care of all your Salesforce needs and concerns.

Feel free to call us at +1 609 632 0350 or write
to us at info@achieva.ai

© 2026 Achieva AI. All rights reserved.